CHARADE: a rule system learning system
作者： Jean-Gabriel Ganascia
DOI:
关键词:
摘要: Designed for an operational prospect, the CHARADE system automatically learns consistent rule systems from a description language, set of axioms reflecting language semantics and examples. The technique advocated below is based on "generate test" mechanism where space explored more general to specific descriptions. Rules properties be obtained are translated into exploration procedure constraints thanks formalization learning with two Boolean lattices.The underlying theoretical framework allows both justify heuristics conventionnaly used similarity based-learning introduce global satisfied by during its construction.
acm.org
ijcai.org 参考文章(4)
Ryszard S. Michalski, A theory and methodology of inductive learning Computer Compacts. ,vol. 1, pp. 49- ,(1983) , 10.1016/0167-7136(83)90132-4
Bruce G. Buchanan, Tom M. Mitchell, Model-directed learning of production rules ACM SIGART Bulletin. ,vol. 63, pp. 44- 44 ,(1977) , 10.1145/1045343.1045371
J. Ross Quinlan, Learning Efficient Classification Procedures and Their Application to Chess End Games Machine Learning. pp. 463- 482 ,(1983) , 10.1007/978-3-662-12405-5_15
Bruce G. Buchanan, Tom M. Mitchell, MODEL-DIRECTED LEARNING OF PRODUCTION RULES1 Pattern-Directed Inference Systems. pp. 297- 312 ,(1978) , 10.1016/B978-0-12-737550-2.50019-1